Trust in Online Collaborative IS - Community-Built Databases: Research and Development - page 342

other similar social software technologies - openness, ease of use, and decentrali-
zation - can also have disruptive consequences on society. Open wikis can easily be
associated with poor quality information and often fall prey to malicious or mis-
leading content editing [ 8 ].
Online communities use trust/reputation management components to facilitate
cooperative user behavior [ 9 ]. In general, trust management systems seek two main
goals: (a) to assist users in rating products or other users for better decision-making,
and (b) to provide an incentive for better user behavior resulting in improved future
performance [ 10 , 11 ]. In the context of wikis, reputation management systems
are suggested as a social rewarding technique that motivates users to participate
actively in sharing knowledge [ 12 ]. In addition, these systems can assist adminis-
trators with automatic detection of high/low reputation users to promote/demote the
access rights.
Reputation can be defined as the public's opinion (more technically, a social
evaluation) of a person, a group of people, or an organization [ 13 ]. Trust is one
user's belief in another user's capabilities, honesty, and reliability based on his/her
own direct experiences. In online communities, there are two notions of trust:
individual-to-individual trust and individual-to-technology trust [ 14 ]. eBay and
online banking are examples of these two categories, respectively. In wikis, we
have a combination of these trust/reputation relationships; individuals need to have
trust in content that is collaboratively created by other individuals. Authors also
need to have trust in other authors collaborating with them to create/edit content.
For example, one of the obstacles faced by experts who collaborate with Wikipedia
is the lack of guarantee that an inexpert/vandal user will not tamper with their
contributed content [ 15 ]. Therefore, the trustworthiness of content is tightly linked
with the reputation of the author.
The remainder of this chapter is as follows: Section 14.2 provides a brief
overview of the relevant literature. Section 14.3 introduces the user reputation
model. Section 14.4 describes how user reputation can be used to assess content
quality. Section 14.5 provides technical background and describes how the data
were collected and mined. Finally, Sect. 14.6 draws some conclusions and points
to a few directions for future investigation.

14.2 Background

Many online communities have trouble motivating enough users to build an active
community. High user participation is the key factor for a successful online
community, and that is why good motivating factors are essential [ 12 ]. As of
April 2010, six of the ten most popular Web sites worldwide simply could not
exist without user-contributed content [ 16 ]. These sites - Myspace, YouTube,
Facebook, eBay, Wikipedia, and Craigslist - look for some incentives to encourage
broader participation or the contribution of higher quality content. In order to
increase and enhance user-generated content contributions,

it is important to